Oenothera Lamarckiana and O. gigas. 969 
relation of the chromosomes when first formed by the segmentation of 
a univalent spireme. 
Summary. 
1. Chromatic bodies, probably representing chromosome centres or 
prochromosomes, may be recognized in the nuclei shortly after the last 
mitoses in the archesporium when the sporophytic chromosomes by alveoli- 
zation have developed delicate nuclear reticula. 
2 . Shortly before synapsis the nuclear reticula become so dense in 
structure through the development and thickening of numerous strands 
that the chromatic bodies can no longer be distinguished, their substance 
apparently merging with that of the deeply staining networks. 
3. The advent of synapsis is indicated by the slow contraction of 
the reticula which carries most of the strands towards the centre of the 
nuclei. During this process the threads become more evident in the 
reticula and come to lie in confused and complicated coils tightly gathered 
together to form the synaptic knot from which a few delicate strands and 
loops extend into the nuclear cavity. These threads appear at times to 
run closely parallel with one another, but the evidence does not justify 
an assumption that this relation indicates more than such a close association 
as would be brought about by the gathering together of a complexly 
looped thread or system of threads. 
4. The synaptic knot, marking the period of mid-synapsis, generally 
lies close to the nucleolus, and is a clearly defined stage extending over 
a considerable period (perhaps one or more days). This is the time in the 
reduction processes of Oenothera when the chromatic elements (prochromo¬ 
somes) are in their most intimate relation to one another. The results 
of the synaptic contraction are an evident thickening of the threads and 
shortening of the thread system. 
5 . Following the period of mid-synapsis, the chromatic material 
emerges from the synaptic knot by a loosening of the coils of the contracted 
thread system. It then becomes evident that the threads have continued 
to thicken and shorten although still complexly looped and coiled. From 
this condition is developed by further condensation the very much thicker 
thread that becomes the univalent spireme. Although irregularities in the 
form and thickness of this spireme may be observed during its development, 
no clear evidence has been noted that it ever becomes doubled by a length¬ 
wise fission to give a ‘ strepsinema * stage. 
6. A segmentation of the spireme gives rise to the full set of sporo¬ 
phytic (somatic) chromosomes arranged end to end. These numbers, 
fourteen in Lamar ckiana and twenty-eight in gigas, are the diploid numbers 
for these species respectively. The spireme segments are always in the 
diploid number, which precludes the possibility of their being bivalent in 
